Unraveling Tumor Evolution
1. Key Concepts: The Language of Cancer Phylogenetics
- Infinite Sites Assumption (ISA): Early models assumed each mutation occurs once. While efficient, this fails to capture loss of mutations (e.g., gene deletions) or recurrent hits in driver genes like TP53 9 .
- Violating ISA: Advanced algorithms (e.g., Dollo/k-j models) allow mutations to disappear and reappear, reflecting real-world complexity like convergent evolution in lung cancer 9 .
- Single-Cell Revolution: Sequencing individual cells reveals subclones—sibling lineages within a tumor. For example, breast CTC clusters contain genetically distinct cells cooperating to metastasize 4 7 .
Single-cell sequencing reveals the hidden diversity within tumors.
2. In-Depth Experiment: The CTC Cluster Detective Story
Objective:
Determine whether circulating tumor cell (CTC) clusters—seedlings of metastasis—arise from one clone (monoclonal) or multiple clones (oligoclonal).
Methodology 4 :
- Sample Collection: Blood from 7 breast/prostate cancer patients + 2 mouse models.
- CTC Isolation: FDA-approved Parsortix microfluidics capture clusters.
- Single-Cell Dissection: Robotic micromanipulation separates cluster cells.
- Whole-Exome Sequencing: Profiles mutations in each cell.
- Phylogenetic Inference: Bayesian model (CTC-SCITE) places cells on evolutionary trees, testing for branching evolution.
Results & Analysis:
- 73% of patient CTC clusters were oligoclonal (Fig 1d).
- Lineage-defining mutations in 40% of breast clusters drove functional divergence (e.g., truncated proteins).
- In mice, oligoclonal clusters surged from 11% (low-diversity tumors) to 68% (high-diversity tumors) (Fig 2b).
Implications:
Oligoclonal clusters act as "cooperative gangs," combining diverse skills to invade distant organs. Disrupting their cohesion (e.g., via Na+/K+ ATPase inhibitors) could block metastasis 4 .
| Cancer Type | Clusters Tested | Oligoclonal (%) | High-Impact Mutations |
|---|---|---|---|
| Breast | 15 | 73% | 40% |
| Prostate | 1 | 100% | 0% |
| Mouse (LM2-NSG) | 11 | 79% | 36% |
3. The Computational Toolbox
Phylogenetic algorithms convert genetic chaos into clear narratives:
- NGPhylogeny.fr: User-friendly platform with "one-click" workflows (e.g., FastTree for large datasets) 2 .
- scPhyloX: Models time-varying dynamics, like stem cell overshoot in fly development or subclonal selection in colorectal cancer 7 .
- reSASC: Simulated annealing approach for mutation loss/reacquisition, outperforming ISA models 9 .
| Reagent/Software | Function | Application Example |
|---|---|---|
| Parsortix | Microfluidic CTC capture | Isolating intact clusters from blood |
| CTC-SCITE | Bayesian phylogenetic inference | Deconvoluting oligoclonal cluster ancestry |
| Lentiviral Barcodes | Track clones in vivo | Quantifying tumor diversity in mouse models |
| scPhyloX | Phylodynamic parameter estimation | Modeling stem cell differentiation |
User-friendly platform for phylogenetic analysis with one-click workflows.
Visit SiteAdvanced tool for modeling time-varying dynamics in single-cell data.
GitHubBayesian method for inferring tumor phylogenies from single-cell data.
4. Clinical Impact: From Trees to Treatment
Evolutionary metrics predict outcomes better than traditional staging:
- In prostate cancer, genomic diversity + Gleason heterogeneity reduced recurrence time by 50% .
- Spatial segregation of clones (HR=2.3) and chromosome 6p loss (immune evasion) were key prognostic markers .
- Phylogenetics guides targeted therapy: HER2-low breast cancers now respond to T-DXd, approved in 2025 8 .
| Biomarker | Hazard Ratio (Recurrence) | Biological Insight |
|---|---|---|
| Genomic Diversity (SNVs) | 3.12 | High intratumor heterogeneity |
| Gleason Morphology | 2.24 | AI-quantified architectural disorder |
| Spatial Clone Segregation | 2.30 | Geographical isolation of resistant clones |
| Chromosome 6p Loss | Immune correlation | Reduced T-cell infiltration |
Phylogenetic analysis enables more precise targeting of therapy-resistant clones, improving outcomes for patients with advanced cancers.